Agentic AI Foundations Associate (1Z0-1157-26) - Practice Exam 1

1. A company hosts several users' LangChain conversations. Which state-management practice is appropriate?
2. A custom tool expects a numeric account identifier, but the model proposes a free-form sentence. What should the tool layer do?
3. A database-grounded assistant gives an answer that conflicts with the retrieved records. Which TWO checks should guide investigation? Select TWO.
4. A developer adds an input guardrail to a customer-facing agent. What role should that control play?
5. A developer calls an agent through a framework's invoke interface. Why might one invocation involve several model and tool interactions?
6. A developer wants orchestration support for agents, tools, handoffs, and execution tracing rather than implementing every loop directly around API calls. Which option is most relevant?
7. A developer wants to compose an agent from a model, tools, instructions, and execution controls using LangChain. What role does LangChain play?
8. A finance agent can issue refunds. Which requirement is most important before enabling this tool for customer-facing use?
9. A hosted agent uses memory to continue tasks across interactions. Which design is appropriate for a multi-user application?
10. A LangChain agent chooses between search_orders and create_refund tools. What most directly helps it understand which tool is appropriate?
11. A model returns a request to call a custom get_invoice function. Who is responsible for actually executing that custom function in the normal function-calling workflow?
12. A purchase-order tool may have succeeded before the network response was lost. What should the team consider before retrying it?
13. A registered tool fails because a downstream service is temporarily unavailable. Which tool behavior best supports a reliable agent?
14. A support assistant must look up an order, inspect the result, and decide whether another tool is needed before answering. What makes this design agentic rather than a single text-generation call?
15. A support system routes billing issues to a specialist billing agent. Which agent-orchestration concept is most relevant when responsibility transfers to that agent?
16. A system hands a task from a general support agent to an account specialist. Which consideration is important when deciding what context to transfer?
17. A team enables tracing for agents that process confidential records. Which TWO practices are appropriate? Select TWO.
18. A team evaluates a tool-using agent before production. Which TWO checks are useful? Select TWO.
19. A team evaluates Oracle AI Database Private Agent Factory. Which use case is most aligned with that capability?
20. A team needs a predictable sequence of three data transformations with no dynamic decisions. Which implementation choice is reasonable?
21. A team needs to investigate an agent that produced the wrong result after several tool calls. Which evidence is especially useful?
22. A team wants agents that operate close to Oracle database data using configured tasks and tools. Which capability should it investigate?
23. A team wants to focus on an enterprise agent's instructions and tools while using a managed runtime for deployment and operation. Which OCI capability should it evaluate?
24. A team's agent is correct but too slow and costly for its workload. What evidence should guide optimization?
25. An agent calls a shipment-status tool and receives a delivery date. What should happen next in the agent loop?
26. An agent discovers a new tool from an MCP server. What must the application still check before allowing a sensitive call?
27. An agent needs to find semantically relevant passages in enterprise documents before answering. Which Oracle AI Database capability is most relevant?
28. An agent repeatedly alternates between two tools without completing its task. Which control should the application enforce?
29. An agent turns a natural-language business question into SQL. What should the application enforce before and during execution?
30. An agent's retrieved document contains a command to ignore all safety instructions. What should the system do with that command?
31. An AI application uses an MCP client to access a server that exposes a customer lookup operation. Which component normally provides the lookup tool?
32. An application executes a model-requested function and receives a result. What should it supply back to the model for continuation?
33. An assistant needs context from earlier messages in the same task. Which mechanism is most directly relevant?
34. An enterprise agent reads internal knowledge content for several departments. What should govern its retrieval permissions?
35. An enterprise agent reports success after a required downstream tool failed. What is the correct reliability requirement?
36. An MCP-enabled client connects to an Autonomous AI Database MCP Server. Which boundary must remain enforced?
37. An MCP integration exposes a read-only company policy document as contextual data. Which MCP primitive is most relevant?
38. An MCP server exposes a reusable instruction template for a common support workflow. Which primitive describes this capability?
39. An organization wants several AI applications to discover and call the same approved enterprise tools through a standard interface. Which problem does MCP help address?
40. Which TWO statements correctly describe the boundaries of MCP? Select TWO.

Comments — 0